This work analyzes the performance of a Large Scale Hydrological Model, which takes into consideration land use changes, in a deforested Amazon basin. Three scenarios were considered in the simulations: the first scenario (LUCC) considers satellite derived land use and land coverage change, from forest to pasture, from 1982 to 2005; the second scenario (NOLUCC) assumed that pristine vegetation remained unchanged during the whole period; and finally the last scenario (CAPO) considers secondary forest growing in deforested areas based on estimations derived using the algorithm of Neef, et al (2007). Secondary vegetation simulations were set to reproduce evaporative fraction behavior observed by Giambelluca et al (2005) in tropical forests of Thailand. Simulations for the first scenario show a good agreement with observed daily discharge at the entire basin and several sub-basins. Despite this, trends in observed discharge series were not captured by the model. Differences between NOLUCC and LUCC simulations suggested an increment in annual water production in the basin for the disturbed condition. When secondary vegetation (CAPO) is considered in the simulations, the impact of LUCC (forest to pasture) in water production was reduced, due to increased evapotranspiration rates. LUCC introduces positive trends in discharges at few sub-basins, meanwhile consideration of secondary forest in the simulations (CAPO) counterbalanced some of those trends. Also, the model was able to simulate seasonal evapotranspiration differences between secondary and primary forest as observed by von Randow, et al (2010).
Redes Sociais